# st: RE: RE: RE: RE: unpaired regression

 From "Scott Merryman" To Subject st: RE: RE: RE: RE: unpaired regression Date Wed, 12 May 2004 19:29:18 -0500

```I have no idea if this is appropriate for your problem, but if you do want
R2 values, how about randomly shuffling the values of assay2 and collecting
the R2 for different combinations.

Something like this:

clear

qui{
set obs 10
gen a1 = uniform()
gen a2 = uniform()

gen r2 = .
gen rn = .
set obs 1000
gen index = _n  in 1/10

forv i = 1/1000 {
gen a2_`i' = .
replace rn = uniform() in 1/10
sort rn
replace a2_`i' = a2[index]
sort index
}

forv i = 1/1000 {
qui reg a1 a2_`i'
replace r2 = e(r2) in `i'
drop a2_`i'
}
}

sum r2

Side Note:

If I combine the above -forv- loop, so it looks like:

clear
qui {
set obs 10
gen a1 = uniform()
gen a2 = uniform()
gen r2 = .
gen rn = .
set obs 1000
gen index = _n  in 1/10
forv i = 1/1000 {
gen a2_`i' = .
replace rn = uniform() in 1/10
sort rn
replace a2_`i' = a2[index]
sort index
reg a1 a2_`i'
replace r2 = e(r2) in `i'
drop  a2_`i'
}
}

sum r2

it only produces around 630 R2 values rather than 1000.  Is there something
I am missing?

Scott

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Wallace, John
Sent: Wednesday, May 12, 2004 10:57 AM
To: 'statalist@hsphsun2.harvard.edu'
Subject: st: RE: RE: RE: unpaired regression

> Can anyone comment on whether Scott's suggestion would be appropriate
> for
> the problem I'm working on?  The difference in R^2 between the samples
> indicates that it might be problematic.

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```